The Agent Skills Directory

[SAFE]: The skill performs intended business research functions without any detected malicious patterns or security violations.
[PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection as it ingests untrusted data from external websites via company URLs.
Ingestion points: SKILL.md (Phase 0 and Phase 1 research steps).
Boundary markers: Absent; there are no explicit instructions to ignore embedded instructions in the fetched web content.
Capability inventory: The skill triggers shell command execution for the tam-builder.py script and hands off context to other lead-finding skills.
Sanitization: No sanitization of external content is specified before it is summarized for the user or passed to downstream scripts.
[COMMAND_EXECUTION]: The skill provides instructions for the agent to execute a local Python script to perform market mapping.
Evidence: python3 skills/capabilities/tam-builder/scripts/tam_builder.py referenced in SKILL.md with various operational flags.
Safety Context: The instructions explicitly require user approval before database writes and recommend using --preview and --sample modes for testing.

icp-identification